A comprehensive study on the quality characteristics of sweet red pepper paste and the evaluation of near-infrared spectroscopy as a rapid alternative tool

Red pepper ( Capsicum annuum L. cv. Capia) is an important vegetable obtained from a 1-year cultivated plant belonging to the Solanaceae family. Among the products of red pepper, its paste is very similar to tomato paste in appearance; however, it differs in terms of flavor properties. The wide-ranging quality parameters (dry matter, soluble solids, pH, titration acidity, salt, insoluble ash, total reducing sugars, color values, total carotenoid, antioxidant properties, browning index, and 5-hydroxymethylfurfural) of commercial sweet pepper pastes ( n = 52) were determined. To the best of our knowledge, this study is the most comprehensive study on determining the quality parameters of red pepper paste, and its in-depth results will help food authorities issue registrations related to commercial pepper paste in countries such as Türkiye. Additionally, near-infrared (NIR) spectroscopy-based machine-learning algorithms were evaluated for the first time in paste samples as a rapid alternative or complementary to the conventional analytical methods. Regarding the conditional entropy-based models, successful predictions with high accuracy were developed for all the quality parameters of pepper paste. Except for the coefficient of determination value of 0.69 and 0.85 in the prediction of HMF and a / b color values in the developed models, respectively, 0.96 or higher coefficients of determination were obtained for all other parameters. Therefore, NIR spectroscopy, along with the conditional entropy-based predictions, enables simple, fast, multiparametric and accurate quantitative prediction of the numerous quality parameters for commercial sweet pepper pastes and has the potential to be used in routine quality control analyses by pepper paste producers.